The website of the Donald Lab at Duke University is an early 2000s relic with Comic Sans font, hypertext menu, and photoshopped cartoons. It does not look like the birthplace of a groundbreaking technology in the fight to cure cancer.
Yet the lab’s 20 years of research into protein design is the foundation of Ten63 Therapeutics, a startup out to drug undruggable cancer targets. Ten63’s platform, BEYOND, can analyze 19,290,123 feasible drug compounds per second in the search for molecules that might inhibit the proteins that drive cancer. Based on a generative quantum chemistry AI, BEYOND not only discovers hidden (cryptic) pockets within previously undruggable targets but also designs drugs for these previously intractable cancer targets.
Like the Donald Lab webpage, Ten63’s PhD co-founders Marcel Frenkel, Mark Hallen, and Bruce Donald, don’t take themselves too seriously. It was a mix of chance, curiosity, and grief that brought them together to create the ambitious suite of algorithms that may one day turn individualized cancer treatment into a routine medical procedure.
In 2014, Marcel Frenkel showed up at a Duke biochemistry retreat. He had a difficult question for every graduate student, postdoc and faculty member who presented, despite being only a second-year grad student himself. Marcel saved his toughest questions for the Donald Lab, which he was already determined to join. Marcel had read papers based on the lab’s research on how OSPREY, a protein design algorithm, could revolutionize the treatment of undruggable diseases including HIV, the subject of his research.
The lab’s founder and namesake, Bruce Donald, was a professor in five different Duke University departments ranging from math to biochemistry and had long ignored the borders of academia. He had earned a PhD in computer science at MIT after studying Russian Language and Literature as an undergraduate at Yale. At Duke in 2006, he launched the Donald Lab to study protein design, which spans applied math, computer science, biochemistry, and chemistry.
One major contributor to Osprey was the computational chemist Mark Hallen. Born in Ithaca, New York, the son of PhDs in physics and electrical engineering, hard sciences were pretty much Mark’s destiny. As a math and chemistry double major at Duke, Mark heard about a lab that had combined math and chemistry in ways that almost sounded like “sci-fi.” It was the Donald Lab, and to Mark’s surprise, it was making progress on protein design. He stayed at Duke for his PhD and joined the lab.

The challenge in protein design is to model trillions of possible protein sequences and find the handful that might work in a specific use case, like inhibiting a drug-resistant tumor. The only efficient way to discover those proteins is through a massive computational effort, hence the Donald Lab’s protein design algorithm, OSPREY.
Mark’s superpower was to make that algorithm work better. “Every year Mark would come up with some strikingly new idea,” recalls Bruce Donald. “He’d read the literature and assimilate it at this very rapid rate and come up with shockingly original takes on it. And then he would implement something that performed better than everything else.” By 2014, OSPREY was being used in clinical trials for biologics and vaccines.
For ten years, Bruce had pitched his PhD students on starting a company around OSPREY. None had stepped up. Marcel would be different. In his suit at the biochem retreat in Durham, Marcel already believed that only with commercial backing could OSPREY reach its full potential to beat cancer. And beating cancer was a very personal mission for him. When he was accepted, Marcel’s main ambition was to launch Osprey outside academia.
Born in San Paulo, Brazil, Marcel grew up in a close knit family accustomed to challenging adversaries. His father was a special agent for the Brazilian IRS specialized in investigating organized and financial crimes. His mother, a lawyer, chose to stay home, partially to help Marcel overcome dyslexia. Thanks to her dedication, says Marcel, he attended the University of Notre Dame as a Hesburgh Scholar and varsity athlete for his undergraduate biochemistry degree.
At Duke, Marcel’s initial research focused on HIV, but in 2014 his mother was diagnosed with pancreatic cancer. The doctors said that nothing could be done. Marcel began to research what technology could treat previously untreatable cancers and, after meeting Bruce at the Duke retreat, applied to the Donald Lab.
“He read probably 20 to 50 papers on his own every day and absorbed them all somehow,” says Bruce. When a friend of Bruce’s was diagnosed with a rare breast cancer, Marcel knew which inhibitors to try and which clinical trial to join. Thanks to Marcel’s advice, Bruce’s friend was soon tumor-free to the shock of her doctors.
“What IndieBio gave us was not only a community of founders, with whom we’ve remained very strongly connected, but also a community of investors. We went from being scientists to entrepreneurs. I don’t think we could have achieved that anywhere else.”
Marcel Frenkel, Co-founder & CEO, Ten63
Bruce credits Marcel with bringing “the lab to the next level.” When the lab needed more computers, Marcel wrote grants to get them. When the lab needed more grad students, Marcel recruited them. When the lab needed to develop new ways to target previously undruggable proteins, he dedicated himself to finding solutions. When Marcel’s mother died in 2015, he and Mark became more determined than ever to shift Osprey from an academic to a clinical setting.
“They did everything you would expect a scientist to do but I think that they were also preparing the technology and the vision for escape,” recalls Bruce. Rather than focus on postdocs or job opportunities in industry, the two talked about how to turn OSPREY into a business.
Marcel officially asked Bruce and Mark to join him in launching this new venture, originally called Gavilán Biodesign, which launched in December 2018. From the beginning, says Marcel, “It was very impact driven. It was about how we can change the lives of patients like my Mom.”

One big technical challenge remained. OSPREY’s algorithms could only be applied to proteins, a very specific albeit vital area. To design next-generation cancer therapeutics, OSPREY’s algorithms had to extend to other chemical spaces and become more accurate. Mark, who was working at the Toyota Technological Institute at Chicago (TTIC), supplied that next foundational piece. “Mark had been working for years on a way to drastically enhance OSPREY’s accuracy and generalize OSPREY’s algorithms to all the chemical space” says Marcel. “Conceptually, this is like jumping from special to general relativity.” Ten63’s BEYOND platform has added nearly 1 million lines of code to OSPREY and extended its applicability into all drug-like chemical space. Today the platform searches proteins and over 100 trillion drug-like molecules with unprecedented accuracy.
Early in 2019, they began applying to startup programs that could teach them to operate like entrepreneurs. They joined SOSV’s IndieBio startup development program in San Francisco, where they booked an over-priced Airbnb in the Haight Ashbury neighborhood of San Francisco, and each day took two buses to get to the IndieBio lab. “It was one of the best moments of our lives,” says Marcel.
At IndieBio, they changed the name to Ten63, which stands for 10 to the 63rd power, the estimated number of compounds with therapeutic properties.

“IndieBio was a 6-month MBA on how to start and build a founder-led company,” recalls Marcel. “What IndieBio gave us was not only a community of founders, with whom we’ve remained very strongly connected, but also a community of investors. We went from being scientists to entrepreneurs. I don’t think we could have achieved that anywhere else.”
When investor meetings and the IndieBio Demo Day in San Francisco arrived in June 2019, Bruce flew in from Duke. He brought a tweed jacket and tie, only to be advised that he should wear jeans and a hoodie. Soon after IndieBio, the company got to work targeting Myc, a protein that is thought to drive 70% of all cancers, including pancreatic cancer.
At present, Ten63 is developing their own internal pipeline and working with pharmaceutical companies, including Boehringer Ingelheim, to discover therapeutics for currently untreatable diseases. The company has raised $21.4 million in venture funding from SOSV, Hatteras Venture Partners, Yosemite, Morpheus Ventures, Draper Associates, and others.

By and large, cancer patients remain confined to traditional chemotherapy, which is highly toxic to the patient. Ten63 believes BEYOND can create such specifically targeted therapeutics that patients will carry on with normal, healthy lives while undergoing cancer treatment.
“Our mission at Ten63,” says Marcel, “is to use next-generation algorithms to discover new therapeutics so that patients have options when facing currently incurable diseases.”
By Richard Ellis
Photos by Amy Stern Photography